Accurate risk stratification has an important role in the management of patients with acute coronary syndromes. Even in patients with ST-elevation acute myocardial infarction (STEMI), for whom early therapeutic options are well defined, risk stratification has an impact on early and late therapeutic decision making. We aimed to compare the prognostic value of 4 risk scores used to evaluate patients with STEMI. We conducted a prospective registry of all patients treated with primary percutaneous coronary intervention for STEMI from January 2001 to June 2006. Excluded were patients with cardiogenic shock. A total of 855 consecutive patients were included in the analysis (age 60.5 +/- 13 years, 19% women, 28% with diabetes, and 48% with anterior wall myocardial infarction). For each patient, the Thrombolysis In Myocardial Infarction (TIMI), Controlled Abciximab and Device Investigation to Lower Late Angioplasty complications (CADILLAC), Primary Angioplasty in Myocardial Infarction (PAMI), and Global Registry for Acute Coronary Events (GRACE) risk scores were calculated using specific clinical variables and angiographic characteristics. Thirty-day and 1-year clinical outcomes were assessed. The predictive accuracy of the 4 risk scores was evaluated using the area under the curve or C statistic method. The CADILLAC, TIMI, and PAMI risk scores all had relatively high predictive accuracy for 30-day and 1-year mortality (C statistic range 0.72 to 0.82), with slight superiority of the CADILLAC score. These 3 risk scores also performed well for prediction of reinfarction at 30 days (C statistic range 0.6 to 0.7). The GRACE score did not perform as well and had low predictive accuracy for mortality (C statistic 0.47). In conclusion, risk stratification of patients with STEMI undergoing primary percutaneous coronary intervention using the CADILLAC, TIMI, or PAMI risk scores provide important prognostic information and enables accurate identification of high-risk patients.